At the moment I cannot test this out, and you'll understand the reason now.
Anyway, I found a possible answer to this problem (and many other errors) in this link: Experts Exchange - Veeam Replication jobs fail with some sort of timeout .
While searching for a solution, I found a funny thing that made me laugh a lot, you can find it here: BigBrus Corner - JOKE: Veeam Client Error Banks Pool Is Suspended
SHORT ANSWER
The link above (This one: Experts Exchange - Veeam Replication jobs fail with some sort of timeout) says that error "Client Error Banks pool is suspended" could be related to network speed between servers. If it's a WAN link, maybe not too fast, fairly used, and you're within a very strongly encrypted VPN, this could lead to a lot of errors like this one, due to rekeying or simply none little bandwidth available.
I'm trying to organize even better than now my backups (so to reduce the amount of data to transfer), to slightly reduce the VPN encryption and to use fastest encryption algorithms. In this way I should increase throughput a little; I'll let you know if it's enough.
The best way to solve this problem would probably be to use a faster link between servers. Since this is not always possible, you should try to increase throughput in some way (reducing encryption, increasing compression, something other that didn't come into my mind).
LONG ANSWER
Keep in mind I'm using Veeam 5 - we'll soon be updating it, but it's quite old.
The real problem regarding Veeam and link problems - either unstable or slow ones - is that they're incredibly difficult to find out. You can find a lot of errors not really useful:
- Client Error Banks pool is suspended
- ReplicateDiskContentSynt failed
Client error: Can not retrieve requested properties.
Soap fault. Detail: '', endpoint: ''
- Server error: End of file
- Client error: The NFC backup does not exists
- Read summary xml error
Root element is missing.
- Failed to delete OIB
Each one of these different errors points to some kind of corruption of the Replica. Once the Replica is corrupted, I didn't find anything anywhere to restore it. The only thing you can do is to delete all of your replication history (for the failed job obviously, not for every job) and start again. The cause of the corruption anyway, is another matter.
What I understood of all this situation is that a replica is a really delicate process. Once started, every problem you can have on the link will corrupt your entire chain of replicas (at least, on my version). Yes, not only this precise replica, the entire chain. And everything Veeam will say to you is some cryptic error like "Client Error: Banks Pool is Suspended". (By the way, what the f**k is a "Bank Pool"???)
On more recent version of veeam, you can configure a WAN accelerator (
Veeam - WAN accelerator), so that you can fully exploit your bandwitdth. Well, at least if you don't have to use it for other scopes - like surfing the internet or serving your customers. Anyway, last I checked, WAN accelerators were available just for Backups and not for Replicas, so in my case it's useless.
As soon as I can tell you for sure that increasing throughput is the way to solve these problem, I'll let you know. I'm still testing at the moment.
Labels: veeam, vmware